14th CONGRESS OF THE INTERNATIONAL SOCIETY OF PHOTOGRAMMETRY
Hamburg 1980
No. of Commission : VII
No. of Working Group: 5
(Invited Paper)
APPLICATIONS OF AIRBORNE THERMAL
INFRARED SCANNERS TO ENGINEERING PROBLEMS
Peter K. Pleitner
Daedalus Enterprises, Inc.
Ann Arbor, Michigan
ABSTRACT
This paper presents an overview of some of the types of engineering problems that have been investigated through the use
of the airborne thermal infrared scanner in the last several
years . The successful use of thermal infrared scanners in
providing data to solve some of these engineering problems
promises to open new prospects for the operational use of
this remote sensing technique as an engineering tool .
Special techniques for the determination of leaks in underground fluid transfer systems and seepage through dams and
dikes are illustrated .
INTRODUCTION
This paper discusses how thermal infrared (TIR) remote sensing technology
can be successfully used by photogrammetrists and civil engineers as a
complementary and unique airborne survey tool . Many valuable applications
provide new data for interpretations and offer new service opportunities
pertinent to civil engineering tasks .
Important principles , operational
considerations, and procedures are reviewed . Most important is the
discussion on how this tool is successfully employed on an operational
basis for a broad variety of survey applications . To evaluate the TIR
survey potential , the application must be defined consisely in light of
the thermal process and properties of the subject . The success of the
thermal survey mission, however, is judged by the interpretability of its
imagery product .
Required here is an understanding of key principles of
detection, what is detected, and proper mission planning . Application
examples of interest to this audience are used to demonstrate how these
factors are integrated for successful TIR survey missions .
PERSPECTIVE
TIR technology development started in the 1950's for the military to detect
the thermal manifestation of matter and display this property relative to
its thermal background . The military ' s interest was the ability of a TIR
scanning system to detect and image remote thermal anomalies and features
camouflaged for low reflected light c ontrast , or hidden by low levels of
reflected light .
Important civilian application for this technology became evident in the
60's when its potential for terrain mapping was recognized . In 1968, the
59£
government declassified systems not exceeding certain thermal and spatial
resolution standards.
The 70's witnessed the gradual transfer of the remote sensing of heat to
scientific and engineering applications. Two significant technological
refinements of military system designs made this transition to civilian
applications viable: 1) The recording of the TIR detector signal on
magnetic tape. The signal could, thereby, be electronically optimized and
manipulated, after acquisition, to generate the desired image qualities.
2) The incorporation of calibrated thermal reference sources.
This
feature transformed a system capable of imaging relative temperature
differences into a radiometrically stable and quantitative imaging
radiometer.
SIMPLE PHYSICS - (Basic Principles)
At absolute zero (0°K or -273°C) all motion of atoms and molecules in
matter is frozen.
As temperature is increased, the random motion of particles produces collisions and accelerations. This internal heat or
kinetic temperature is converted to radiant energy.
All materials at a temperature above absolute zero radiate energy in a
characteristic and quantitative manner that depend on the temperature and
emissivity of the material.
Emissivity (s) represents the efficiency of a
material's surface in radiating energy.
Our skin will sense radiated heat
that is emitted energy traveling as electromagnetic radiation (EMR). We
call this energy thermal infrared and define it to be that region of the
electromagnetic spectrum with wavelengths (A) longer than reflected infrared and visible EMR, but shorter A than microwave and radio EMR, Figure 1.
We refer to a specific region or band of TIR by identifying its A range.
By convention A is measured in terms of micrometers (1 ~m
1 x 10-6m).
EMR may be transmitted, absorbed, emitted, scattered, and reflected.
The
interaction between EMR depends on atomic, molecular, size, and surface
property of matter.
Smoother surfaces have lower emissivity values than
rough surfaces of the same composition. Many atmospheric gases such as
ozone, water vapor, and carbon dioxide absorb radiation in the TIR region.
Figure 1 shows two TIR wavelength regions where the atmosphere is most
transparent to TIR.
These are called TIR transmission windows.
The 3 to
5 ~m window represents the shortest TIR A range in common use.
Hot matter (450°C) radiates more energy in this region than at longer A.
If heat or molecular action is increased further, it will become visible
to the eye or influence photographic emulsion. The second window, from
8 to 14 ~m, is the region of peak radiation of matter at ambient earth
temperature. This window is the A region which offers the greatest
variety of TIR remote sensing applications potential.
Heat travels from one place to another by three mechanisms:
conduction,
convection, and radiation.
Solar radiation is the principle source of
thermal energy radiated from the ground.
Solar radiation varies in duration and intensity depending on the time of day and season.
Figure 2
illustrates a typical heating and cooling cycle of a variety of surface
cover types.
Local topographic orientations modify this diurnal effect.
Many mathematical relationships describe the behavior of EMR. Of the
various equations that are used in thermal radiation calculations, the
following are most important.
597
Wien's Displacement Law : The wavelength of maximum emittance is proportional to a physical constant divided by its radiant temperature in degrees
Kelvin .
A max
2897 . 9 ]Jill (°K)
T (°K)
Stefan-Boltzmann Law : The total energy emitted by a blackbody is proportional to the fourth power of its absolute temperature .
A blackbody (BB) is a theoretical material
absorber of TIR . No energy is reflected .
BB is perfect or 1 . Emissivity represents
surface to radiate energy. No material is
"graybodies" have values between 0 and 1.
which is a perfect emitter and
The emissivity coefficient of a
the efficiency of a material ' s
a perfect BB . The emissivity of
Emitted energy in the TIR A range is related to four internal properties of
materials .
A.
Thermal Conductivity - The measure of the rate at which heat
will pass through a material ; conduction is the transfer of
heat through molecular action .
cal
cm • sec • °C
B.
Thermal Capacity - The ability of a material to store
heat .
cal
gram · °C
C.
Thermal Inertia - A measure of the thermal response of the
cal
material to temperature changes .
D.
Thermal Diffusivity - The rate at which a temperature can
be transferred between the surface and interior of a
material.
(Sabins, 19 78)
Refer to Table 1 for data on geologic materials .
Most natural materials have fairly high emissivities (coefficients close
to 1) in the TIR region. Kirchhoff ' s Law states the ratio of radiant
emittance to absorptance is the same for all bodies. At thermal equilibrium, emissivity is equal to absorptance; therefore, high emissivity indicates high absorptance. It follows that detected thermal induced radiation
is basically a surficial phenomenon. Conditions below the surface (density, conduction, convection, heat flow, moisture content, etc.), as well
as atmospheric conditions such as wind, humidity, solar radiation, etc . ,
affect the detected temperature of the surface .
THE TECHNOLOGY
To detect a given object or surface by thermal infrared means, it must
have a radiation characteristic that differs from its surroundings, i.e . ,
a different temperature or emissivity . TIR line scanners are devices
capable of sensing this thermal radiation . Emitted temperature differences of 0 . 2°C are routinely detectable. Special purpose systems can
598
discriminate differences measured in hundredths of a degree.
TIR at the
most useful wavelengths cannot pass through glass and cannot be recorded
directly by film emulsion.
Thus to detect and image this EMR, reflective
optics and a quantum detector, analogous to one silver halide crystal, are
employed in airborne systems. The detector has an electrical output which
changes proportional to minute variations of the incoming TIR radiation.
A rotating mirror allows the detector to look at the ground scene in a
sweeping motion perpendicular to the flight line.
Its electrical output
is amplified and recorded on magnetic tape or recorded directly on a moving
film strip via a CRT tube trace of each scan line on the ground .
A gyroscope provides a roll stabilizing signal for the start of each line. The
forward progression of the aircraft and film speed adjustment provides the
contiguous scan lines of the image. The final product is a photo-like
image . The pattern of gray tones, which may be color-coded, represent
radiant temperature-emissivity differences rather than reflected
differences from a surface.
By convention,
light tones represent warm areas of higher radiation levels, and dark
tones represent cool areas on photographic print. Quantitative scanners
compare the incoming TIR signal relative to two precisely calibrated and
stable thermal reference sources.
They are adjusted to bracket the temperature range of interest.
Each scan line the TIR signal is compared to
the references; therefore, each signal level, gray tone, or color represents the same radiation levels, or temperature, throughout the image.
The diagrams of Figure 3 show the important geometric relationships and
operational parameters of a typical line scanner system.
PROBLEM DEFINITION AND APPLICATION EVALUATION
Assessing the potential for successful applications of TIR remote sensing
techniques requires a concise definition of thermal and spatial features
to be imaged relative to thermal processes, operational conditions, and
TIR principles.
Direct applications are those where the detection, quantification, and mapping of heat distribution are of interest .
Indirect
applications are those where the apparent radiation temperature of a
surface, relative to that of its surroundings, will be used as an indicator
of conditions related to the survey objective.
Examples of the former are
fire detection through smoke or below ground, heat loss, and measurement of
thermal effluents; of the latter are geotechnical applications where relative heat capacity and/or conductivity can be used to interpret density,
moisture content, and water dynamics as indicators of subsurface conditions.
In both cases, the correct interpretation of the measurements
depends upon knowledge of the physical characteristics of the material and
the understanding that the thermal effect imaged and measured is a property
of the surface only, but is influenced by internal and external parameters.
Examples of TIR applications to engineering problems are used to demonstrate how successful TIR surveys are conducted.
HEAT LOSS SURVEYS
DATA ACQUISITION - The objective of a heat loss survey is to obtain an
image which can be interpreted with confidence to assess the heat retention
efficiency of a structure. A structure is subject to internal and external
thermal loads. To image heat lost from the inside, the cumulative effect
of solar heating must be allowed to reradiate before meaningful TIR data
acquisition .
599
Line scanner geometry, aircraft altitude, and recording medium determine
imagery scale.
Up to 4X enlargement of a film record can be made without
loss of spatial detail. Ambient ground temperature should be 2°C or less
(with lower dew point) for best thermal contrast. Sky conditions must be
clear or uniformly overcast; surface winds not exceeding 10 km/hour. High
wind will cause "smear" along the boundaries between features of different
radiation temperature.
Trees, snow cover, and standing water will attenuate or absorb thermal emission. The best time of acquisition is between
several hours after sunset to predawn.
A daytime aerial photo taken within
one day of a TIR survey is helpful to the client and provides the interpreter with a visual comparison standard for interpretation of the thermal
TIR data.
Imagery interpretation must be based on the following relationships and
principles:
A.
Spatial resolution on the surface is determined by the altitude
of the aircraft and each apparent temperature is the average
value of one IFOV.
B.
The recorded radiation temperature is a function of all of the
physical parameters discussed above as modified by the intervening atmosphere.
C.
Metalic surfaces reflect the thermal irradiance of the night
sky, approximately - 60°C on a clear night, therefore, appear
black.
D.
Heat is dissipated by radiation, conduction, convection, and
ventilation, so that the TIR survey measures only heat lost
by radiation.
SEEPAGE THROUGH DAMS & DIKES - The TIR survey objective is to detect
anomalous soil moisture conditions and water seepage on the face of an
earthen retention structure and in front of its toe . Standing water is
usually represented as a light tone on a nighttime image, due to its high
heat capacity as compared to surrounding soil, rock, or vegetation materials. The thermal contrast of flowing water depends on its temperature
difference relative to adjoining surficial material.
The thermal contrast
of moist soils depends on the rate of evaporation and its heat capacity.
Damp ground is usually represented as darker gray tone, relative to
adjoining dry soil during the day and has the opposite appearance at night
due to the much higher heat capacity of wet soils compared to dry soils .
The predawn TIR image and daytime photo comparison of Ill. 1 shows clearly
the merit of thermal detection of seeps and soil moisture vs. tonal contrast due to reflected light. This imagery was acquired predawn from 600m
AGL.
Ground temperature was -2°C with 6Km/hr surface wind and a clear sky.
TIR data were recorded on magnetic tape for laboratory playback and imagery production . The continuous tone analog data displays wet areas and
probable seeps as light tonal features or warm, relative to dry soils.
Thermal reference sources of the line scanner were used to calibrate and
level slice the imagery into eight tonal levels between 24°F and 36°F.
The six gray levels represent 2°F temperature increments . This quantification of radiant temperature permits the analysis of soil temperature
across the face of the dam and below the dam .
Finer sets of isolevels were
produced bracketing only the temperature variations of the seepage areas.
This allowed a comparison of seep temperature to know ground water springs,
600
the surface temperature of the impoundment, and the discharge ; thereby,
interpretation progressed to probable seep sources .
BURIED PIPELINE LEAK DETECTION - Breaks in underground steam or fluid
transfer pipelines are detected using TIR scanner imagery on an operational
basis. Many have been subsequently verified by excavation . Heat or moistur e will produce thermal anomalies on the ground . Their combined affect
will be imaged, given the proper ground and survey conditions .
Buried pipeline systems present considerable variability in terms of line
size , proximity of lines, topographic position, type, and depth of cover .
Imagery should be interpreted using information on line size, spacing,
insulation characteristics, backfill material, depth of burial , and line
temperature information.
GEOTECHNICAL AND SITING SURVEYS
TIR surveys in areas with sparse vegetation and thin soil cover provide
optimum conditions for obtaining detailed imagery of near surface bedrock
structure. Imagery of areas covered by thin soils underlayed by faulted
and fractured bedrock or a limestone/dolomite sequence display structural
features which can be interpreted faster and cheaper using TIR remote
sensing than by any other survey method . Major linears, which were not
resolved on reflected light and near IR imagery, have been imaged with TIR
scanner systems even in temperate
areas. 1 Some such linears have been
verified to be faults by seismic methods and proposed nuclear power plant
sites were relocated .
The Florida Department of Transportation 2 has demonstrated the successful
application of nighttime TIR imagery interpretation to locate solution
features and sink holes. Over 70% of the anomalous thermal areas interpreted to be probable sink holes were verified by subsequent borings.
Investigators in South Africa 3 have published very impressive and conclusive results for geotechnical studies based upon TIR line scanner imagery
interpretations . Sinkholes, pinnacles, faults, fractures, joints, potential water well sites, borrow pit areas, and impoundment site suitability
studies have been successfully interpreted from TIR imagery of many localities underlayed by near surface dolomite . They stress the importance of
much additional information obtainable from TIR imagery and the complementary benefit of combining thermal geologic interpretation with photogeological interpretations . Typically they perform TIR surveys predawn between
300m and 3000m altitude . The majority of missions are flown in the dry
season when vegetation cover is sparse. Surveys are not conducted immediately after rainfall . However, judicious timing after precipitation may
enhance interpretation of soil permeability, potential aggregate reserves,
and ground water regimes .
Illustration 2 shows a spectacular application, personal communications,
and (op. cit.) capable of saving vast sums of money and potentially lives
was demonstrated in South Africa by chance . The imagery was analyzed after
acquisition only for equipment related R&D purposes . However, it clearly
showed two anomalous dark-tone high moisture bands on the median of a
divided highway cut on a steep slope . Six months after acquisition, the
slope failed precisely along the most prominent thermal linear . Remedial
measures (possibly dewatering and slope toe abutment) along the second
anamaly have reportedly been undertaken .
601
Figure 1.
BLACK
Block body ol 5800 °K
t
BODY RADIATION
and
SUN'S RADIATION
CURVES
energy
>-
l!l
0:
body al 1200°K
UJ
z
Black body al 600°K
body ot 300 °K
I 1111
40 60
100
UJ
I
lm
of
I
I
10m lOOm
eorlh's
atmosphere
SPECTRAL RANGE OF OPERATION
FOR COMMON REMOTE SENSING INSTRUMENTS
Radar
"0
1+----- Human Eye
c:
0
Q)
Photography
c: c c
c
0
0
~
XOd..
·f------->-j
~
f-
"0 "0 "0
Q) Q) Q)
~r---I-M_u_lt~i_-~S~p~e~c~lr_o_I~S~c_o~n~n~e~rs~--~
j...
.2
.4
The electromagnetic energy spectrum
(~rom
Sherz and Stevens, 1970).
Figure 3 .
R
Scan Mirror Rotation
Ground Speed
UJ
FOV (Total Field of View)
.O.UJ = Spatial Resolution
in mrad. (str. rad.)
H
Altitude
D
Scan Line Spacing
B
Swath Width
= 2 [H TANW/2]
V
- - ---v
VV
TFOV (Inst. Field of
View = mrad. x H)
Film is S-bend Corrected
& Rolls Stabalized
Continuous Film Strip
-
.........
v -
1~,"~'~
Determined by V/ H
Scale is H dependent
602
Table l
Thermal properti es of geologic ma terials and water at 20°C
Geologic
materials
K
c
Thermal
conductivity,
cal · em - 1
• sec-1 , oc - 1
p
Density,
gm • cm- 3
Thermal
capacity,
cal· gm - 1
• oc-1
p
k
Thermal
Thermal
inertia ,
diffusivity,
cal· em - 2
1
cm 2 ·sec - 1 • sec- 12 • oc- 1
Basalt
0.0050
2.8
0.20
0.009
0.053
Clay soil (moist)
0.0030
1. 7
0.35
0.005
0.042
Dolomite
0.0120
2.6
0 .1 8
0.026
0 .075
Gabbro
0 .0060
3 .0
0.17
0.012
0.055
Granite
0.0075
0.0065
2 .6
0.16
0.016
0.052
Gravel
0.0030
2.0
0.18
0.008
0.033
Limestone
0 .0048
2 .5
0 .17
0.011
0 .045
Rhyolite
Sandy gravel
0.0055
0.0060
2 .5
2.1
0 .16
0.20
0.014
0.014
0 .047
0 .050
Sandy soil
0.0014
1.8
0.24
0.003
0 .024
Sandstone, quartz
0.0120
0.0062
2 .5
0.19
0.013
0.054
Serpentine
0 .0063
0 .0072
2 .4
0 .23
0.013
0 .063
Shale
0.0042
0.0030
2.3
0.17
0.008
0.034
Tuff, welded
0.0028
1.8
0.20
0.008
0.032
Water
0 .0013
1.0
1.01
0.001
0.037
*Source: From Janza (1975 , Table 4 . 1).
Emissivity of representative samples
of various materials determined in the
8 to 12 pm wavelength region
Material
Emissivity,
Granite
0.815
Feldspar
0 .870
Granite, rough
0.898
Quartz sand , large grains
0.914
Dolomite, rough
0.958
Asphalt paving
0 .959
Concrete walkway
0.966
Water, with a thin film of petroleum
0 .972
Water, pure
0.993
*Source: Buettner, K. J. K., and C. D. Kern. Journal of
Geophysical Research, v. 70, p. 1333, 1965, copyrighted
by American Geophysical Union.
603
f
In the 8 to 12 ~m region,
polished metal surfaces have very
low emissivities of about 0 . 06,
but a coat of flat black paint
increases the emissivity to
about 0. 9 7 .*
It is important to note these
are laboratory measurements. In
the field, E is averaged over one
IFOV by a line scanner. Similar
E ranges could be measured by
soil types and grain size.
*Adapted From Sabins, 1978
X X X X X X X X X X X X X X X X A X X X
HEAT ED SOURCES
SUNSET
Figure 2.
DIURNAL THERMAL CYCLE
J
L>
z
<
~
,-- -r ----- r ·- - .- --.----1
0001
0400
OBOO
MIDNIGHT
Illustration 1.
Daedalus Enterprises, Inc .
604
12 00
NO ON
1600
2000
2400 HOURS
MIDNIGHT
Illustr at i on 2 - Courtesy of Spectral Africa (Pty . ) Ltd.; R. P. Vilj oen and T. D. Feuchtwanger
m
0
Ul
CONCENTRATION
SYNOPSIS OF WIDEBAND TIR IMAGERY APPLICATIONS
Forest Fire Surveys, Smoke and Cover Penetration
Volcanic Activity and Geothermal Surveys
Coal Refuse Dump and Insite Coal Fire Surveys
Power Line Fault Surveys
Process Temperature Surveys and Refineries
Heat Loss from Industrial and Urban Areas, Urban Analysis
Microclimate and Airmass Dynamics Studies
Detection of Ancient Cultural Features, Archeology
Geotechnical Studies, Siting, Routing, and Hazard Surveys
Sink Hole Detection, Karst Topography Surveys
Near Surface Aggregate Surveys
Fault and Fissure Surveys, Moisture Anomaly Delineation
Slope Failure Susceptibility Studies
Dam and Dike Integrity Surveys
Pipeline Surveys, Buried and Exposed
Crop Moisture, Stress and Irrigation Studies
Soil Moisture Aquifer Recharge Area Definition
Ground Water Discharge to Surface and Water Bodies
Thermal and Sewage Effluent Discharge to Water Bodies
Tidal and Current Studies
Detection and Delineation of Oil Spills
Ice, Permafrost, Glacial, and Snow Covered Coast Line Studies
Census and Behavior Studies of Hidden Mammals
1Daedalus Enterprises, Inc.
2 Florida
Department of Transportation, Tech Memo No. 1, 1972, and Tech
Memo No. 5, 1972, by Tom Griepentrog.
3Spectral Africa (Pty.) Ltd., D. Warwick, P. G. Hartopp, R. P. Viljoen,
Q.Jl Engrg. Geol., 1979, Vol. 12, pp. 159-179.
BIBLIOGRAPHY
Griepentrog, T., (1972); Technical Memorandums No. 1, January, 1972, and
No. 5, June, 1972: Florida Department of Transportation.
Sabins, F. F., (1978); "Remote Sensing: Principles and Interpretations":
W. H. Freeman & Company, San Francisco, pp. 426.
Schertz, J. P., and Stevens, A. R., (1970); "An Introduction To Remote
Sensing For Environmental Monitoring": Madison, Wisconsin, The
University of Wisconsin, pp. 80 .
Warwick, D., Hartopp, P. G., and Viljoen, R. P., (1979); "Application of
the Thermal Infra--red Linescanning Technique to Engineering Geological
Mapping in South Africa": Quarterly Journal by Engineering Geology;
Vol. 12, pp. 159-179.
606